DESCRIPTION (Investigator's Abstract): This competing continuation application continues our efforts to prevent post-traumatic seizures and builds on our 10 years experience in evaluating phenytoin (Dilantin Prophylaxis of Post-traumatic Seizures) and valproate (valproate Prophylaxis of Post-traumatic Seizures). The first study (years 01-06) showed that phenytoin prevents early seizures but is ineffective in preventing late post-traumatic seizures. The second study (years 07-1I) was designed to test the hypothesis that valproate would prophylax against post-traumatic seizures and is presently progressing well. It is now in its fourth year with 269 patients randomly assigned to blinded treatment and 24 months of follow-up. This application requests a study extension for both the accession (6 months) and follow-up (24 months) sufficient to achieve study goals. The present study examines the effectiveness of valproate in preventing post-traumatic seizures using a randomized, double-blind design. The aims of the study are: I) to determine whether valproate and phenytoin differ in their effectiveness for preventing early post-traumatic seizures; 2) to determine whether valproate and phenytoin differ in their ability to prevent late post- traumatic seizures or postpone their onset; 3) to determine whether 6-months of treatment with valproate prevents seizures better than I month of treatment; 4) to evaluate the medical cognitive emotional. and psychosocial side effects of valproate. Patients with severe head injuries with at least a 20% chance of developing late post-traumatic seizures are the population of interest. Patients with preexisting conditions that could cause seizures or exacerbate potentially-dangerous side effects are excluded. Patients receive the loading dose of valproate or phenytoin within 24 hours of injury, and then subsequent doses to maintain serum levels within the therapeutic range. The control regimen consists of 1 week of phenytoin followed by placebo. The experimental regimen consists of valproate for 1 or 6 months. Patients take blinded medication for 6 months and are- followed for an additional 18 months without drugs. Incidence of post- traumatic seizures is determined over 2 years. Comprehensive cognitive, effective and psychosocial evaluations are conducted at 1, 6, and 1 2 months after injury. Because of a delay caused by the FDA and by somewhat lower accession than anticipated, the study cannot provide conclusive answers to the questions posed in the time originally proposed. The requested continuation will allow definitive determination of the effects of valproate on post- traumatic seizures and on the patient functioning.